G’day BRD, > On 1 Jun 2016, at 3:33 PM, Bernard Robertson-Dunn <[email protected]> wrote: > > If you look at flying air planes, why are they not all fully automatic? > Flying a plane is a far simpler problem. The reasons why planes haven't > been fully automated may well apply to cars. And cars have their own, > additional problems.
Actually, commercial aeroplanes have pretty much been fully automated. You can enter coordinates/destinations/flight plans into the modern autopilot and the plane will take off, fly to the destination and land without any human intervention at all. I’m not too sure about the taxiing, but everything about the flying can effectively occur without any intervention from the pilot. A number of airlines got themselves into trouble with this - putting people behind the control column who were effectively the world’s worst (but cheapest) pilots. The range of conditions and emergencies that autopilots can handle also increases on an annual basis. That’s not to say that autopilots don’t have their limitations … sensors may fail, hardware may fail, software may hit a set of conditions nobody imagined … but MANY more accidents occur when the human pilot doesn’t trust the instruments and sensors, when external pilot feedback fails due to clouds, storms, night conditions or low light, when rough weather/microblasts occur too fast and too erratically for a human to react to, etc. etc. > > I'd be far happier if the objective was to develop cars with an > automatic mode that could be engaged under specific circumstances (like > an advanced cruise control). Across the board, all cars being driverless > seems a stretch too far. Too much Jetsons. Jetsons is probably 50 years away … as I said, the road to transport automation will be a gradual one - if for no other reason that we simply don’t have either the money or resources to implement any all encompassing ‘solution’ in one fell swoop. We’re talking an infrastructure and technology investment of hundreds of billions (just for Australia) if we wanted to do it that way. > >> Bottom line, there are risks with both ‘systems’ … but the automated one is >> more likely to be fine tuned and perfected much more easily. The one that >> relies on humans, with all their inherent imperfections, will continue to >> carry all the attendant risks. I’ve seen nothing in my life that indicates >> that the human race will improve … but technology does demonstrably improve. > > Sometimes. Sometimes it makes things worse. And it can get worse faster. True. > And changing technology can be much harder than changing human behaviour > - given the right incentives. I did Honours level Psychology, amongst other things, during my academic career and would dispute that statement. First, only a Behaviouralist would agree with you, and I’d suggest they would be agreeing with you without much evidence. Secondly, many human mental ‘pathologies’ and emotions (recklessness, frustration, fear, anger, grief, guilt, panic, excitement or even simple joie de vis etc etc) and behaviours that affect driving capabilities and performance are either regarded as so common they don’t need to be treated, or are regarded as so endemic to the human condition as to not need treatment. Third, for those of the standover/threat school of adjusting human behaviour, we have in place a huge body of Law and a police force that’s meant to make driving safe, but accidents still occur, traffic jams and congestion still happen, and the Courts are full of an increasing incidence of serious driver and traffic offences - even after years of same. (Do we need to go back to the bloke running ahead of the car with a flag to reduce it?) As for technology. well, seat belts came in for a reason - and did what they were mandated to do. I still don’t like them, but I can’t deny their effectiveness. Air bags have stopped numerous casualties. The collapsing-front-end design of modern cars makes them less lethal, ABS lets you brake more controllably, independent suspension makes your ride more comfortable. The rear mounted cameras are - probably (I’ve seen no figures yet) - reducing accidents. The same will probably occur with all the other (prospective) incremental changes that we’ve already covered. Technology will improve driving, improve survivability, improve traffic and other vehicle handling capabilities. Nothing I’ve seen indicates that humans are on anything like that safety/performance improvement curve. In fact, quite the opposite. Removing humans from the equation can only be an improvement. (That said, I’ve been a grumpy old anthropomorphic bastard for years … so my views on automation may be coloured by my dislike of my fellow Man.) Just my 2 cents worth ... _______________________________________________ Link mailing list [email protected] http://mailman.anu.edu.au/mailman/listinfo/link
